Hindawi Biochemistry Research International Volume 2018, Article ID 6406372, 9 pages https://doi.org/10.1155/2018/6406372

Research Article Postrecruitment Function of Yeast Med6 Protein during the Transcriptional Activation by Mediator Complex

Gwang Sik Kim and Young Chul Lee

School of Biological Science and Technology, Hormone Research Center, Chonnam National University, Gwangju, Republic of Korea

Correspondence should be addressed to Young Chul Lee; [email protected]

Received 20 September 2017; Accepted 29 October 2017; Published 9 January 2018

Academic Editor: Andrei Surguchov

Copyright © 2018 Gwang Sik Kim and Young Chul Lee. ,is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Med6 protein (Med6p) is a hallmark component of evolutionarily conserved Mediator complexes, and the genuine role of Med6p in Mediator functions remains elusive. For the functional analysis of Saccharomyces cerevisiae Med6p (scMed6p), we generated a series of scMed6p mutants harboring a small internal deletion. Genetic analysis of these mutants revealed that three regions (amino acids 33–42 (Δ2), 125–134 (Δ5), and 157–166 (Δ6)) of scMed6p are required for cell viability and are located at highly conserved regions of Med6 homologs. Notably, the Med6p-Δ2 mutant was barely detectable in whole-cell extracts and purified Mediator, suggesting a loss of Mediator association and concurrent rapid degradation. Consistent with this, the recombinant forms of Med6p having these mutations partially (Δ2) restore or fail (Δ5 and Δ6) to restore in vitro transcriptional defects caused by temperature-sensitive med6 mutation. In an artificial recruitment assay, Mediator containing a LexA-fused wild-type Med6p or Med6p-Δ5 was recruited to the lexA operator region with TBP and activated reporter gene expression. However, the recruitment of Mediator containing LexA- Med6p-Δ6 to lexA operator region resulted in neither TBP recruitment nor reporter gene expression. ,is result demonstrates a pivotal role of Med6p in the postrecruitment function of Mediator, which is essential for transcriptional activation by Mediator.

1. Introduction species, placing this evolutionarily conserved coregulator as an essential and general player in Pol II transcription [5, 6]. Promoter-specific mRNA synthesis requires a minimal set of Yeast Mediator is required for diverse aspects of tran- proteins comprising Pol II (RNA polymerase II) and asso- scriptional regulation, such as activation, repression, stim- ciated GTFs (general transcription factors), which is defined ulation of basal transcription, and TFIIH-dependent as in vitro “basal transcription” [1]. Regulated responses of phosphorylation of the CTD (C-terminal repeat domain) of this minimal transcription system by DNA-binding tran- the Rpb1 subunit in vitro [2, 4]. ,e tight association of scription factors or in vitro transcription from nucleosomal Mediator with the CTD of Pol II forms Pol II holoenzyme, DNA template require additional sets of coregulator pro- which is generally required for the transcription of most teins involved in chromatin remodeling/modification as well class II genes in vivo [7]. Genetic and biochemical studies as targeted recruitment of basal transcription machinery to identified 21 polypeptides as bona fide subunits of the core the promoter [2, 3]. In vivo, these transcriptional cofactors Mediator, including transcriptional regulators previously indeed participate in the regulated expression of specific genes identified by genetic studies, as well as the novel subunits in response to environmental and physiological cues, which collectively named Med proteins [8]. In addition, a subset of has been demonstrated in a wide range of eukaryotic model Srb/Ssn proteins was identified as the distinct kinase module organisms. whose association with core Mediator is known to suppress Mediator complex was first identified from yeast the positive function of Mediator [9, 10]. S. cerevisiae [4], and Mediator homologs or its related Molecular and genetic studies have revealed that some complexes were subsequently found in other eukaryotic Mediator subunits are specifically required for the regulation 2 Biochemistry Research International of a subset of genes, whereas others are necessary for general and relay a regulatory signal from the tail/middle module to transcription of Pol II [11]. Consistent with these observa- the head module/Pol II as proposed [12]. tions, our biochemical studies have demonstrated that Although extensive structural studies have been made to functionally related Mediator subunits physically associate reveal the action mechanisms of Mediator in Pol II tran- to form two stable subcomplexes, named Rgr1/Med14 and scription, analyses of each Mediator component to confirm Srb4/Med17 [12]. ,e Med14 subcomplex was subsequently the functions inferred from structural data remain to be shown to contain several functional modules, including completed. In order to enhance our understanding of the key Gal11/Med15 module (Med2, -3, -15, and -16) and Med9/10 role of S. cerevisiae Med6p (scMed6p) in Mediator functions, module (Med1, -4, -7, -9, -10, and -21), which suggested we generated a series of Med6p mutants harboring small a modular structure of Mediator [12, 13]. While the Med14 internal deletions (10 amino acids) and analyzed the im- subcomplex consists of several functional modules pre- portance of each region of scMed6p in transcription acti- sumed to function in receiving diverse signals from acti- vation by Mediator in vivo and in vitro. We identified three vators or repressors, the Med17 subcomplex, composed of distinct domains of scMed6p required for its essential Med6p and genetically dominant Srb proteins (Med17, -18, function: one N-terminal region (Δ2, amino acids 33–42) -20, and -22), is thought to functionally interact with the might be involved in Mediator association and/or protein Rpb1 CTD [12]. Based on electron microscopic (EM) im- stability, and the second (Δ5, amino acids 125–134) and third ages, the modular structure of core Mediator was repre- (Δ6, amino acids 157–166) regions have a role in post- sented as comprising distinct head, middle, and tail domains recruitment function, such as signal-relay or conformational [14]. Recent X-ray structures of yeast Mediators suggest that changes during transcriptional activation by Mediator. the Med17 subcomplex corresponds to the head module while the Med14 subcomplex contains both tail and middle 2. Materials and Methods modules [15–18]. In holoenzyme and preinitiation complex (PIC) models based on refined EM and X-ray structures, 2.1. Construction of med6 Deletions, Mutants, and Other both the middle and head modules make multiple contacts ExpressionPlasmids. Serial deletion mutants of the scMED6 with diverse parts of Pol II and GTFs, explaining the mo- gene were constructed by an oligonucleotide-directed lecular mechanism of Mediator-stimulated CTD phos- mutagenesis method. ,e MED6 gene harboring own phorylation by TFIIH and resultant enhancement of basal promoter and terminator regions was obtained from transcription [15, 16, 19]. pRS316-MED6 by EcoRI-XbaI digestion and subcloned Med6p is a hallmark component of evolutionarily into the pALTER-1 vector (Promega) for site-directed conserved Mediator complexes and participates in tran- mutagenesis according to the manufacturer’s instruc- scriptional activation of many class II genes in S. cerevisiae tions. A series of mutagenic oligonucleotides were designed [5, 20]. In metazoans, mutant flies deficient in dMED6 to delete 10 amino acids of Med6p spanning the whole exhibit transcriptional defects in a wide range of develop- MED6 ORF region (Figure 1). A NarI restriction site was mentally regulated genes, validating the functional conser- introduced into deletion points within each mutagenic vation and the requirement for Med6p in Pol II transcription oligonucleotide, which was used for the construction of in vivo [21]. Intriguingly, med17 mutation was isolated as double deletion mutants (Δ3-4, Δ7-8, and Δ9-10). After the a dominant suppressor of med6 temperature-sensitive (ts) deletions were confirmed by sequencing, mutant med6 mutation [12], and this genetic interaction was confirmed genes were subcloned into pRS313 plasmid. by an opposite genetic approach [22]. Consistent with For the copper-inducible expression of Med6p, wild- functional interaction, a direct physical association between type and mutant MED6 genes were subcloned into pC Med6p and Med17p was identified through successful in (native form) or pCL (LexA-fused form) vectors using ap- vitro reconstitution of the Med17 subcomplex [13, 23]. propriate restriction sites [26]. pGEX-MED6 was made by Recently, X-ray structural data have been obtained for the subcloning the EcoRI-SalI fragment from pGBT-MED6 into head modules of S. cerevisiae and S. pombe Mediators, which pGEX-2TK (Pharmacia). ,e pGEX-med6 deletion con- indicate that the structures of two head modules are well structs were made by replacing the indicated region of conserved despite the limited sequence similarities between pGEX-MED6 with the corresponding fragments of mutant these regions [17, 18]. ,e conserved Med6p structure constructs in pRS313 by the use of restriction sites. In order revealed a core domain, which comprises a five-stranded to make the E. coli expression constructs for 6xHis-tagged antiparallel β-sheet, two pairs of α-helices flanking a con- Med6p and its deletion derivatives, the EcoRI-ClaI frag- served groove, and a flexible C-terminal α5 helix [18]. ,is ments from the pGEX-MED6 and deletion constructs were domain is called “shoulder,” according to its position relative subcloned into pET vector series. to the head module, and was shown to interact with other subunits within the head module (Med8 and Med17) and the middle module (Med4, Med10, and Med19) [18]. Recently, 2.2. Growth Complementation Assay. To test the growth the C-terminal α6 helix of the Med6 subunit was shown to complementation by med6 deletion mutants, yeast strain connect to the Med14 subunit of the middle module as one YCL4 (MATα, ade2, his3, trp1, lys2, med6Δ::LEU2, and of the tethering molecules of the head module [15, 16]. ,ese pRS316-MED6) was used as a host strain for plasmid results strongly suggest that the Med6p may function as shuffling. ,e med6 deletion mutants in pRS313 were in- a critical interface between the head and middle modules troduced into YCL4 strain and the resulting transformants Biochemistry Research International 3

spMed6

scMed6 spMed6 cMED6 dMed6 hMED6

spMed6

scMed6 spMed6 cMED6 dMed6 hMED6

spMed6

scMed6 spMed6 cMED6 dMed6 hMED6

spMed6

scMed6 spMed6 cMED6 dMed6 hMED6

scMed6 spMed6 cMED6 dMed6 hMED6

spMed6

scMed6 spMed6 cMED6 dMed6 hMED6

Figure 1: Multiple-sequence alignments of Med6p homologs. Sequences of Med6 proteins from S. cerevisiae (scMed6), S. pombe (spMed6), C. elegans (cMED6), D. melanogaster (dMed6), and H. sapiens (hMED6) are aligned with the aid of Expresso (Di Tommaso et al. [24]) and presented by ESPript 3.0 program (Robert and Gouet [25]) along with the secondary structure information of spMed6 (pdb code 4H63 and 5N9J). Black and shaded boxes indicate identical and similar amino acids, respectively, and gaps are represented as dots. Ten regions chosen for the internal deletion (10 amino acids each) are also shown as thick lines just below the corresponding amino acid sequences of scMed6p. ,e mutational sites found in med6-ts1 and med6-ts2 alleles are depicted as sharp (#) and asterisk (∗), respectively. were tested for their growth on synthetic complete media 2.3. Protein Purification. Gal4-VP16, GTFs, wild-type, and containing 5-fluoroorotic acid via the plasmid shuffling med6 ts mutant Pol II holoenzymes were prepared using method. several chromatographic steps as described previously [20]. 4 Biochemistry Research International

Recombinant Med6p and mutant derivatives were expressed then transformed into host strain YCL4 for plasmid shuf- in E. coli strain BL21(DE) (Novagen) and purified through fling and tested for supporting cell viability upon the re- Ni+-NTA columns (Qiagen) under denaturing conditions. moval of wild-type MED6 gene in pRS316 by 5-fluoroorotic Purified proteins were renatured at 4°C through stepwise acid treatment (Figure 3(a)). Despite the considerable size dialysis according to the protocol described by ,ompson for deletions (10 amino acids), seven out of ten deletion et al. [27] and stored at −80°C before use. mutations (Δ1, Δ3, Δ4, Δ7, Δ8, Δ9, and Δ10) had no obvious effect on the viability of yeast cells (Figures 2 and 3(b)). Notably, the med6 deletion mutants containing Δ2 (residues 2.4. In Vitro Transcription. Reconstituted in vitro tran- 33–42), Δ5 (residues 125–134), or Δ6 (residues 157–166) scription was performed as described previously [20]. Either were unable to support the cell growth. wild-type or med6 ts mutant Pol II holoenzymes were ° ° ,e multiple-sequence alignment of Med6p homologs preincubated for 10 min at 25 C or 37 C with other sup- using the Expresso program indicated a high degree of plements containing two DNA templates (pGAL4:G- and conservation between yeast and metazoan Med6 proteins pGCN4:G-), GTFs, 0.5 mM ATP, and Gal4-VP16, with or (Figure 1). However, scMed6p has distinct features from without recombinant Med6p derivatives. After initiation other Med6 homologs in four regions: two internal and one complex formation, the transcription reaction (30 min, ° 32 C-terminal spacer regions (50, 8, and 30 amino acids long, 25 C) was initiated by the addition of [α- P]-UTP and CTP. resp.) that are exclusively found in scMed6p and one variable ,e purified reaction products were analyzed on the 7% region among Med6 homologs (amino acids 230–240 in denaturing PAGE and autoradiography. scMed6p). According to alignment data, three regions (Δ2, Δ5, and Δ6), identified as essential for MED6 function(s), 2.5. LacZ Reporter Assay, GST Pull-Down Analysis, Western closely matched the highly conserved regions of Med6 Analysis, and Chromatin Immunoprecipitation. For the homologs (Figure 1). analysis of reporter gene expression in yeast, yeast strains To confirm the functional regions of scMed6p, larger were grown to the mid-log phase in selective synthetic deletion mutants were made and used in a complementation complete media and β-galactosidase activity of cultured cells test. First, we constructed a Med6p mutant having the was measured in triplicate by the permeabilized-cell method combined deletion of Δ3 and Δ4 (Δ3-4), resulting in the [28]. To investigate the expression levels of mutant Med6ps, removal of the first region specific for scMed6p. Interestingly, yeast strain YCL4 was transformed with plasmids encoding this mutant could fully complement the MED6 function, LexA-Med6p derivatives (in pCL313) and/or native forms of indicating that this region is dispensable for essential function Med6p derivatives (in pC314). ,e resulting transformants (Figures 2 and 3(b)). In addition, one-third of the C-terminal were cultured in synthetic glucose media containing 0.5 mM domain of scMed6p, corresponding to the most divergent CuSO4 for 2 h to induce the expression of Med6 proteins. region of Med6p, also proved to be unnecessary for MED6 Preparation of yeast whole-cell extracts (WCEs), western function, since two mutants (Δ7-8 and Δ9-10) spanning analysis, and immunoprecipitation of Mediator were per- this region had no obvious effect on cell survival (Figures 2 formed according to the protocols described previously [12]. and 3(b)). ,is assumption was confirmed by the fact that For chromatin immunoprecipitation, L40 yeast strains scMed6p comprising only residues 1–210 could support cell expressing LexA-Med6p derivatives (in pCL313) were grown viability (see Supplementary Figure 1). in synthetic media containing glucose (Glc) or galactose plus Taken together, the functional domains of yMed6p es- raffinose (Gal + Raf) for 6 h in the presence of 0.5 mM sential for cell viability were mapped to the highly conserved CuSO4. Cells were harvested, fixed in 1% formaldehyde regions of Med6p, whereas the scMed6p-specific region and solution, and sonicated for the fragmentation of chromatin the divergent region of Med6ps were not required for es- according to the protocol described by Kuo and Allis (1999). sential function(s) of scMed6p. WCEs containing chromatin fragments were immunopre- cipitated with the use of appropriate antibodies and sub- jected to a decrosslinking reaction. Purified DNAs were 3.2. Phenotypes or Transcriptional Defects of the Viable Med6 subjected to PCR (25 cycles) with the use of oligomers Mutants. Although the deletion analysis indicated that specific to promoter regions of LacZ reporter or GAL1-10 substantial parts of scMed6p were dispensable for cell genes, respectively. viability, a significant level of sequence conservation is apparent at these regions, including Δ1 and Δ7 regions. 3. Results ,us, we investigated whether the viable med6 deletion mutants showed any growth defects or transcriptional de- 3.1. Identification of yMed6p Regions Required for Cell fects in vivo, such as ts lethality or limited carbon source Viability. For the unbiased and systematic analysis of utilization, as observed in med6 ts mutants [12]. All of the yMed6p, a series of Med6p mutants harboring a small in- viable mutant strains had no observable ts phenotype and ternal deletion was made. We designed 10 amino acid de- utilized galactose or raffinose at 30°C as efficiently as the letions throughout the entire Med6p for approximately wild-type did (Supplementary Figure 2A). every 30 amino acid interval and introduced these into Based on the fact that MED6 is essential for galactose- MED6 genes harboring own promoter and terminator re- induced GAL1 gene activation [20], β-galactosidase activity gions (Figures 1 and 2). ,e mutant med6 constructs were of the GAL1UAS-TATA-LacZ reporter was measured for the Biochemistry Research International 5

β 1 In vitro -gal 1 50 100 150 200 250 295 Viability activity2 activity3 WT Med6 ++++++ Δ1 (2–11) + Δ2 (33–42) −−(+) Δ3 (64–73) + Δ4 (95–104) + Δ5 (125–134) − − ++++ Δ6 (157–166) − − (+) Δ7 (189–198) + + Δ8 (221–230) + Δ9 (254–263) + Δ10 (286–295) + Δ3-4 (64–104) + Δ7-8 (189–230) + Δ9-10 (254–295) + Figure 2: Schematic presentations of Med6p derivatives and summary of their functional defects in vivo and in vitro. Viability1 was tested for the ability of deletion mutants to complement the med6 null mutation via plasmid shuffling as described in Materials and Methods. In vitro activity2 represents the capability of each recombinant form of Med6 proteins to rescue the in vitro transcriptional defects of med6-ts2 holoenzyme based on the quantitative data shown in Figure 4. β-galactosidase activity3 indicates the expression levels of the reporter gene in the artificial recruitment assay of the indicated Med6p derivatives as shown in Figure 6(a).

function(s) of scMed6p resides in the evolutionarily con-

med6::LEU2 served regions.

med6Δ MED6 3.3. In Vitro Transcriptional Activity of Mutant Med6 HIS3 URA3 Proteins. Next, we investigated whether scMed6p regions essential for viability are also required for Med6p-dependent transcriptional activation in vitro. For this, we purified recombinant Med6p having ts-2 mutation or internal de- (a) letion mutations and examined their capabilities in restoring Δ3-4 Δ6 (−) the transcriptional defects of Pol II holoenzyme caused Δ9-10 by med6 ts-2 mutation [20]. When transcriptional PIC Δ7 WT was formed at 25°C, both wild-type and med6-ts2 Pol II Δ5 Δ7-8 holoenzymes showed 23- and 21-fold activation of GAL4 enhancer-containing template (GAL4:G-) by Gal4-VP16 Δ1 over basal transcription from GCN4:G-template, re- Δ8 spectively (Figure 4, lanes 1 and 2). However, PIC formation Δ4 (−) at 37°C specifically impaired the transcription activity of Δ2 med6-ts2 holoenzyme in that the mutant holoenzyme gave Δ9 Δ3 Δ10 only a 2.5-fold activation in comparison to the 21-fold ac- (b) tivation by wild-type holoenzyme (Figure 4, lanes 3 and 4). Figure 3: Identification of the regions in Med6p required for cell ,is temperature-dependent transcriptional defect of mu- viability. (a) Schematic depiction of plasmid shuffle technique for tant holoenzyme was recovered by the addition of the the complementation test of mutant MED6 genes (med6Δ). (b) ,e recombinant form of wild-type Med6p, but not Med6-ts2 growth complementation test of the med6 deletion mutants. Wild- protein, prior to the PIC formation at 37°C (lanes 5–7). Next, type (WT) MED6 and indicated med6 deletion mutants in the we examined whether each deletion mutant of Med6p (-Δ2, pRS313 vector were introduced into the host strain YCL4 (med6Δ:: -Δ5, -Δ6, and -Δ7) had the ability to restore the tran- LEU2 and pRS316-MED6) for plasmid shuffling. ,e resulting scriptional defect of med6-ts2 holoenzyme. ,e Med6p-Δ7, transformants were grown for 4 days at 30°C on synthetic complete which showed no apparent defect in vivo, did rescue the media containing 5-fluoroorotic acid to remove the pRS316-MED6. activation defect of the heat-inactivated med6-ts2 holoen- (–): pRS313 vector only. zyme (lane 11). However, Med6p-Δ5 and -Δ6 mutants were not able to complement this defect (lanes 9 and 10), in good viable mutants under induction conditions. None of the agreement with their in vivo phenotypes. Notably, Med6p- viable med6 deletion mutants were defective in transcrip- Δ2, which did not support cell viability, could partially tional activation of the GAL1 promoter (Supplementary restore the transcriptional defect of med6-ts2 holoenzyme to Figure 2B). ,ese results strongly suggest that the pivotal half that of the wild-type (lanes 5, 6, and 8; see Discussion). 6 Biochemistry Research International

PIC formation at 25ºC37ºC pC-MED6 − WT WT Δ2 Δ5 Δ6 Pol II holoenzyme WT Ts WT Ts pCL-MED6 WT Δ2 Δ5 Δ6WTΔ2 Δ5 Δ6 + r- Med6p WT Ts Δ2 Δ5 Δ6 Δ7 α-LexA pGCN4:G-

pGAL4:G- α-Med6

α-TBP

123 4 5 6 78 91011 1234 5678 9101112 Transcription (%) — — 100 12 69 59 18 38 10 15 50 α-Med14 I.P with α-Med14 Figure 4: ,ree lethal Med6p mutants also show transcriptional α-LexA defects in a reconstituted in vitro system. Either wild-type (WT) or 13 14 15 16 med6 ts-2 mutant (Ts) Pol II holoenzyme (700 ng each) was pre- Figure 5: ,e deficiency of Med6p-Δ2 in WCEs and purified ° ° incubated for 10 min at 25 C or 37 C with other supplements Mediator fraction. (lanes 1–12) WCEs were prepared from YCL4 containing two DNA templates, GTFs, 0.5 mM ATP, and Gal4- yeast strains expressing the indicated LexA-Med6p derivatives VP16 (30 ng), with or without the indicated recombinant Med6p (pCL-MED6) along with (+) or without (−) coexpression of the (80 ng each). After PIC formation, the transcription reaction ° 32 native forms of Med6p derivatives (pC-MED6) by copper in- (30 min, 25 C) was initiated by the addition of [α- P]-UTP duction for 2 h. Samples were resolved on an 8% denaturing and CTP. ,e mRNA products were purified and resolved on polyacrylamide gel, and expression levels of LexA-fused and native 7% denaturing polyacrylamide gel, followed by autoradiogram. forms of Med6p derivatives were examined by western analysis Transcription (%) was set to 100% as to the fold activation using anti-LexA antibody and anti-Med6 antibody, respectively. (signal from GAL4 template divided by signal from GCN4 Western analysis for TBP was done for loading control. ,e template) shown by wild-type holoenzyme upon PIC forma- ° degradation product of LexA-Med6p-Δ2 is indicated by an asterisk. tion at 37 C. ,e data quantitation was performed with the use (lanes 13–16) WCEs were prepared from the indicated yeast strains of Personal Molecular Imager FX system and associated and subjected to DEAE column fractionation to enrich the Pol II program (Bio-Rad). holoenzyme. Pol II holoenzyme in each sample was immuno- purified with beads-coupled anti-Med14 antibody and subjected to ,ese results indicate that the Med6p regions which are western analysis to measure the amounts of LexA-Med6p de- essential for in vivo function (cell viability) are also required rivatives (α-LexA) contained in the immunopurified Mediator for activation function of Med6p in vitro. fraction (α-Med14).

in immunopurified Mediator as compared to wild-type 3.4. Med6p-Δ2 Has a Defect in Association with Mediator. In and other mutant Med6ps (lanes 13–16). ,is observa- our previous reports, the Pol II holoenzyme prepared tion was reproduced when WCEs were prepared from from med6-ts mutants is deficient in the med6-ts protein yeast strains expressing LexA-fused wild-type Med6p but the free form of med6-ts protein was not detected and native forms of Med6p mutants (Figure 5, lanes during column purifications, possibly due to its rapid 9–12). In this case, we could clearly observe the re- degradation [12, 20]. ,is result suggests that all the ciprocal expression pattern between LexA-fused and Med6p mutants detected in WCEs appear to be associated native forms of Med6ps, indicating their competitive with Mediator in vivo. Since the malfunction of three incorporation into Mediator. From these results, we Med6p mutants (Δ2, Δ5, and Δ6) could be reasoned from concluded that Med6p-Δ2 has a defect in Mediator as- disabled interaction with Mediator, we investigated the sociation and the Δ2 region (amino acids 33–42) of expression levels of LexA-fused forms of Med6p mutants scMed6p might be involved in the direct interaction with in WCEs in the presence of wild-type Med6p expressed by other Mediator subunit(s). own promoter. Upon copper induction, the expression level of LexA-Med6p-Δ5 and -Δ6 mutants was about half that of the wild-type, whereas the Δ2 mutant was nearly 3.5. Artificial Recruitment Assay with the Functionally absent in the prepared WCEs (Figure 5, lanes 1–4; Figure Defective Med6p Mutants. ,e artificial recruitment of 6(a)). ,e additional copper induction of the native form Mediator to a promoter region via Mediator subunit of wild-type Med6p resulted in a similar pattern, in- fused with DNA-binding domain resulted in activator- dicating that overexpression of wild-type Med6p does not bypass activation of target promoter. We successfully per- affect the expression pattern of these mutants (Figure 5, formed an artificial recruitment assay by using Mediator lanes 5–8). Notably, LexA-Med6p-Δ2 showed a degra- containing LexA-Med6p [29]. ,us, we used LexA-fused dation pattern as indicated by the detection of only LexA- Med6p mutants for the artificial recruitment assay to acti- sized protein, demonstrating the absence of Med6p-Δ2 in vate the chromosomal 8xlexAop-LacZ reporter gene. As the cell lysate was not due to gene expression (lanes 2 and 6). expected from the expression data, LexA-Med6p-Δ2 was Consistent with this, LexA-Med6-Δ2p was totally absent completely inactive for reporter gene activation (Figure 6(a)). Biochemistry Research International 7

300 Glc Gal + Raf LexA-Med6 WT WT Δ2 Δ5 Δ6 GAL1-10p Input 200 8xlexAop

GAL1-10p α-LexA 100 8xlexAop -Gal activity (A.U) β -Gal activity GAL1-10p α-Med14 8xlexAop 0 Δ Δ Δ LexA-Med6 WT 2 5 6 GAL1-10p α-TBP 8xlexAop α-LexA 12345 (a) (b)

Figure 6: Artificial recruitment assay and chromatin immunoprecipitation analysis of Med6 mutants. (a) Reporter gene activation by targeted recruitment of Mediator via LexA-fused Med6p derivatives. Yeast strain L40 harboring pCL313-Med6 derivatives were cultured in synthetic glucose media containing 0.5 mM copper. ,e β-galactosidase activity of chromosomal 8xlexAop-LacZ reporter gene in cultured cells was measured in triplicate. (b) Chromatin immunoprecipitation for recruitment of Mediator (α-Med14), LexA-Med6ps (α-LexA), and TBP (α-TBP) to promoter regions of chromosomal reporter (8xlexAop) and GAL1-10 (GAL1-10p) genes. L40 strains expressing indicated pCL313-Med6 derivatives were grown in synthetic media containing glucose (Glc) or galactose plus raffinose (Gal + Raf) for 6 h in the presence of copper. Chromatin fragments were prepared from WCEs and immunoprecipitated with the use of the indicated antibodies. DNAs were purified from immunoprecipitates and subjected to PCR (25 cycles) with the use of oligomers specific to promoter regions of reporter gene (PCR product, 200 bp) or GAL1-10 gene (PCR product, 370 bp), respectively. ,e amount of DNA used for INPUT PCR corresponds to 1% of the immunoprecipitated DNA samples used in PCR.

Intriguingly, LexA-Med6p-Δ5 had comparable transcrip- demonstrate a pivotal role of Med6p in postrecruitment tional activity to that of wild-type Med6p, suggesting its function (i.e., signal-relay or conformational changes) of transcriptional defect can be reversed by artificial recruit- Mediator during transcriptional activation, even though the ment via an unknown mechanism (see Discussion). In the Δ5 and Δ6 regions of Med6p might have distinct roles as case of the Med6p-Δ6 mutant, its artificial recruitment had suggested by the different responses in the artificial re- no effect on reporter gene expression (Figure 6(a)). cruitment assay (Figure 6(a); see Discussion). Next, we performed chromatin immunoprecipitation analysis to examine in vivo association of LexA-Med6p de- 4. Discussion rivatives, Med14 (for Mediator), and TBP (for transcription initiation) with a target promoter region (8xlexAop). As shown In this report, we dissected the functional domains of yeast in Figure 6(b), Mediator containing LexA-fused Med6p-Δ5 or Med6p on the basis of its requirement for cell viability and -Δ6 was recruited to the 8xlexAop region with an efficiency transcriptional activation via molecular genetics and bio- similar to that of wild-type (irrespective of growth condi- chemical approaches. ,e functional domains of scMed6p tions). However, the association of TBP with the promoter (Δ2, Δ5, and Δ6 regions) identified by complementation region was observed only in the strains expressing LexA-fused assays were mapped to the highly conserved regions of wild-type Med6p or Med6p-Δ6, in agreement with reporter Med6ps, whereas yeast-specific (spacer region) and mostly gene expression (Figure 6). divergent regions (one third of C-terminus) of scMed6p Additionally, we examined the recruitment of Med6p were not required for its essential function(s) (Figures 1 mutants to an endogenous GAL1-10 promoter region under and 2). ,ree lethal deletion mutants had defects in tran- either repression (glucose) or induction conditions (galac- scriptional activation in a reconstituted in vitro system al- tose plus raffinose) to monitor the effect of med6 mutations though only the Gal4-VP16 activator was tested in our in on the recruitment of Mediator to a natural promoter. LexA- vitro transcription (Figure 4). ,ese results again clearly fused Med6p-Δ5 and -Δ6 were recruited to the GAL1-10 demonstrate that the major and essential function of Med6p promoter as well as wild-type Med6p upon galactose in- is focused on transcriptional activation. duction, indicating these mutations did not impair Mediator One interesting point is that some residual activity to recruitment in the endogenous promoter (Figure 6(b)). support in vitro transcription was retained in Med6p-Δ2, In other words, in vivo defects of Med6p-Δ5 and -Δ6 did showing a discrepancy with its in vivo defect. Although we not result from a problem in Mediator-recruitment step. cannot explain clearly this phenomenon, we surmise that Promoter association of TBP (α-TBP) was also seen in all this partial activity might have resulted from an in vitro cases, as predicted by the action of native form of Mediator dosage effect. If Med6p-Δ2 has a normal activation function containing endogenous wild-type Med6p for the activation but has a defect in Mediator association, Med6p-Δ2 could of GAL1-10 promoter (Figure 6(b)). All these results directly provide in vitro activation function to some extent via 8 Biochemistry Research International a dosage effect when an excess amount of Med6p-Δ2 is Med6p but is not documented to interact directly with other added (which is not applicable in vivo). Mediator subunit(s) [18]. In contrast to the Δ5 mutant, the In contrast to Med6p-Δ5 and -Δ6 mutants, the Med6p-Δ2 Med6p-Δ6 protein was completely inactive in the tran- protein was barely detectable in WCEs and was even absent in scriptional activation when artificially recruited to target the enriched Pol II-Mediator fraction. As mentioned above, the reporter (Figure 6(a)). ,e Δ6 region of scMed6p spans β5 comparable proportions of med6-ts proteins detected in sheet and N-terminal part of α5 helix based on X-ray WCEs seem to be associated with Mediator in vivo. ,e structure [18]. Since α5 helix makes critical contacts between observed deficiency and the degradation pattern of Med6p-Δ2 Med8 and Med17 subunits within the head module, the in WCEs suggest that the Δ2 region of scMed6p might be deletion of Δ6 region might cause the defective interactions involved in the association with other Mediator subunit(s) of Med6p with these subunits. According to this scenario, (Figure 5). Based on recent X-ray structure of S. pombe the Med6p-Δ6 protein cannot relay the activation signal Mediator in PIC, the C-terminal residues of α3 helix of from the tail/middle module to the head module, which is spMed6p are shown to directly interact with the Med10/Nut2 not rescued by an artificial recruitment as previously shown subunit to form a head-middle module interface [16]. Since by Med18/Srb5-deficient holoenzyme [29]. the Δ2 region of scMed6p contains N-terminal residues of α3 Collectively, all these results consistently support a notion helix, its deletion might disrupt Med6-Med10 interaction by that Med6p has a pivotal role in the postrecruitment function affecting α3 helix structure (Figure 1). In this scenario, the of Mediator during transcriptional activation, which is con- dissociated form of the Δ2 mutant seems to be rapidly re- ceptually accepted as the conformational changes, isomeri- moved via specific degradation pathway probably due to its zation, or modulation of Pol II activity. In addition, our results instability. Notably, one (Q49L) of three critical mutations in also suggest there might be functional differentiation between the med6-ts2 allele is found in the α3 helix of scMed6p, the Δ5 and Δ6 regions (involved in the relay of upstream suggesting the molecular basis of the deficiency of med6-ts2 versus downstream signal). Future studies on the detailed protein in WCEs and purified Mediator [20]. ,e other two functions of these essential domains of Med6p should be mutations (I68L and L94P) were mapped to the Δ3 and Δ4 helpful for the understanding of action mechanisms of Me- regions, respectively, which correspond to the nonessential diator with regard to its structural and functional relationship. spacer region of scMed6p (Figure 1). Previous biochemical analyses and artificial recruitment 5. Conclusion experiments using mutant Pol II holoenzymes suggested that facilitated recruitment of a holoenzyme to a promoter is ,rough the systematic dissection of functional domains of necessary but not sufficient for transcriptional activation and scMed6p, we identified three distinct domains of scMed6p that an unknown biochemical activity is required for tran- required for its essential function: one N-terminal region scriptional activation (postrecruitment step) [29]. Consistent (Δ2, amino acids 33–42) might be involved in Mediator with this, the structural data suggested that Med6p may association and/or protein stability, and the second (Δ5, function as a critical interface between the head and middle amino acids 125–134) and third (Δ6, amino acids 157–166) modules, and relay a regulatory signal from the tail/middle regions have a role in postrecruitment activity of Mediator module to the head module/Pol II [15–17]. ,is signal relay complex, such as signal-relay or conformational changes. seems to be achieved via multiple interactions of Med6p with ,is study proved the previous notion that Med6p is in- other subunits: Med14 in the tail/middle module; Med4, volved in the postrecruitment function of Mediator com- Med10, and Med19 in the middle module; and Med8 and plex, which is required for gene activation by Mediator. Med17 in the head module [15, 16]. Our artificial recruitment assay revealed that the tran- Abbreviations scriptional defects of Med6p-Δ5 and -Δ6 in vivo did not CTD: C-terminal repeat domain result from a problem in the Mediator-recruitment step GTFs: General transcription factors (Figure 6(b)), which is mainly achieved by an activator-tail Pol II: RNA polymerase II module interaction. In this regard, the defects of Med6p-Δ5 ts: Temperature-sensitive and -Δ6 mutants in activated transcription might reflect PIC: Preinitiation complex their impairments in activation signal relay and inabilities to TBP: TATA-binding protein accomplish intermediary functions that are essential for cell WCEs: Whole-cell extracts. viability. Intriguingly, the transcriptional defects of Med6p- Δ5 seem to be rescued by an artificial recruitment (Figure 6). ,is activator-bypass activation was observed in an artificial Conflicts of Interest recruitment experiment using a holoenzyme devoid of the ,e authors declare that there are no conflicts of interest tail module, which is defective in the Mediator-recruitment regarding the publication of this article. step [29]. From this, we proposed that in vivo defect of Med6p-Δ5 could be the result of a flaw in the reception of Acknowledgments upstream activation signal from the tail/middle module, which can be rescued by an artificial recruitment. According ,is research was supported by Basic Science Research to the structural data, Δ5 region, corresponding to β3 sheet Program through the National Research Foundation of Korea of scMed6p (Figure 1), is critical to form a core domain of (NRF) (NRF-2017R1D1A1B03034080) to Young Chul Lee. Biochemistry Research International 9

Supplementary Materials [15] K. L. Tsai, X. Yu, S. Gopalan et al., “Mediator structure and rearrangements required for holoenzyme formation,” Nature, Supplementary Figure 1: ,e growth complementation test of vol. 544, no. 7649, pp. 196–201, 2017. Med6p comprising residues 1–210. Supplementary Figure 2: [16] K. Nozawa, T. R. Schneider, and P. Cramer, “Core Mediator ,e viable Med6 mutants showed no detectable phenotypes or structure at 3.4 A˚ extends model of transcription initiation transcriptional defects. (Supplementary Materials) complex,” Nature, vol. 545, no. 7653, pp. 248–251, 2017. [17] T. Imasaki, G. Calero, G. Cai et al., “Architecture of the Mediator head module,” Nature, vol. 475, no. 7355, References pp. 240–243, 2011. [18] L. Larivi`ere, C. Plaschka, M. Seizl, L. Wenzeck, F. Kurth, [1] M. Hampsey, “Molecular genetics of the RNA polymerase II and P. Cramer, “Structure of the Mediator head module,” general transcriptional machinery,” Microbiology and Mo- Nature, vol. 492, no. 7429, pp. 448–451, 2012. lecular Biology Reviews, vol. 62, no. 2, pp. 465–503, 1998. [19] P. J. Robinson, M. J. Trnka, D. A. Bushnell et al., “Structure of [2] L. C. Myers and R. D. Kornberg, “Mediator of transcriptional a complete Mediator-RNA polymerase II pre-initiation regulation,” Annual Review of Biochemistry, vol. 69, complex,” Cell, vol. 166, no. 6, pp. 1411–1422.e16, 2016. pp. 729–749, 2000. [20] Y. C. Lee, S. Min, B. S. Gim, and Y. J. Kim, “A transcriptional [3] G. J. Narlikar, H. Y. Fan, and R. E. Kingston, “Cooperation mediator protein that is required for activation of many RNA between complexes that regulate chromatin structure and polymerase II promoters and is conserved from yeast to transcription,” Cell, vol. 108, no. 4, pp. 475–487, 2002. humans,” Molecular and Cellular Biology, vol. 17, no. 8, [4] Y. J. Kim, S. Bj¨orklund, Y. Li, M. H. Sayre, pp. 4622–4632, 1997. and R. D. Kornberg, “A multiprotein mediator of tran- [21] B. S. Gim, J. M. Park, J. H. Yoon, C. Kang, and Y. J. Kim, scriptional activation and its interaction with the C-terminal “Drosophila Med6 is required for elevated expression of repeat domain of RNA polymerase II,” Cell, vol. 77, no. 4, a large but distinct set of developmentally regulated genes,” pp. 599–608, 1994. Molecular and Cellular Biology, vol. 21, no. 15, pp. 5242–5255, [5] S. Malik and R. G. Roeder, “Transcriptional regulation 2001. through Mediator-like coactivators in yeast and metazoan [22] T. I. Lee, J. J. Wyrick, S. S. Koh, E. G. Jennings, E. L. Gadbois, cells,” Trends in Biochemical Sciences, vol. 25, no. 6, and R. A. Young, “Interplay of positive and negative regu- lators in transcription initiation by RNA polymerase II ho- pp. 277–283, 2000. loenzyme,” Molecular and Cellular Biology, vol. 18, no. 8, [6] M. Boube, L. Joulia, D. L. Cribbs, and H. M. Bourbon, pp. 4455–4462, 1998. “Evidence for a mediator of RNA polymerase II transcrip- [23] S. S. Koh, A. Z. Ansari, M. Ptashne, and R. A. Young, “An tional regulation conserved from yeast to man,” Cell, vol. 110, activator target in the RNA polymerase II holoenzyme,” no. 2, pp. 143–151, 2002. Molecular Cell, vol. 1, no. 6, pp. 895–904, 1998. [7] C. M. ,ompson and R. A. Young, “General requirement for [24] P. Di Tommaso, S. Moretti, I. Xenarios et al., “T-Coffee: a web RNA polymerase II holoenzymes in vivo,” Proceedings of the server for the multiple sequence alignment of protein and National Academy of Sciences of the United States of America, RNA sequences using structural information and homology vol. 92, no. 10, pp. 4587–4590, 1995. extension,” Nucleic Acids Research, vol. 39, pp. W13–W17, [8] S. Bj¨orklundand C. M. Gustafsson, “,e yeast Mediator 2011. complex and its regulation,” Trends in Biochemical Sciences, [25] X. Robert and P. Gouet, “Deciphering key features in protein vol. 30, no. 5, pp. 240–244, 2005. structures with the new ENDscript server,” Nucleic Acids [9] T. Borggrefe, R. Davis, A. Bareket-Samish, Research, vol. 42, pp. W320–W324, 2014. and R. D. Kornberg, “Quantitation of the RNA polymerase II [26] D. H. Kim, G. S. Kim, C. H. Yun, and Y. C. Lee, “Functional transcription machinery in yeast,” Journal of Biological conservation of the glutamine-rich domains of yeast Gal11 Chemistry, vol. 276, pp. 47150–47153, 2001. and human SRC-1 in the transactivation of glucocorticoid [10] H. Spahr,˚ O. Khorosjutina, V. Baraznenok et al., “Mediator receptor Tau 1 in Saccharomyces cerevisiae,” Molecular and influences Schizosaccharomyces pombe RNA polymerase II- Cellular Biology, vol. 28, no. 3, pp. 913–925, 2008. dependent transcription in vitro,” Journal of Biological [27] C. M. ,ompson, A. J. Koleske, D. M. Chao, and R. A. Young, Chemistry, vol. 278, pp. 51301–51306, 2003. “A multisubunit complex associated with the RNA poly- [11] S. J. Han, Y. C. Lee, B. S. Gim et al., “Activator-specific re- merase II CTD and TATA-binding protein in yeast,” Cell, quirement of yeast mediator proteins for RNA polymerase II vol. 73, pp. 1361–1375, 1993. transcriptional activation,” Molecular and Cellular Biology, [28] L. Guarente, “Yeast promoters and lacZ fusions designed to vol. 19, no. 2, pp. 979–988, 1999. study expression of cloned genes in yeast,” Methods in En- [12] Y. C. Lee and Y. J. Kim, “Requirement for a functional in- zymology, vol. 101, pp. 181–191, 1983. teraction between mediator components Med6 and Srb4 in [29] Y. C. Lee, J. M. Park, S. Min, S. J. Han, and Y. J. Kim, “An RNA polymerase II transcription,” Molecular and Cellular activator binding module of yeast RNA polymerase II ho- Biology, vol. 18, no. 9, pp. 5364–5370, 1998. loenzyme,” Molecular and Cellular Biology, vol. 19, no. 4, [13] J. S. Kang, S. H. Kim, M. S. Hwang, S. J. Han, Y. C. Lee, pp. 2967–2976, 1999. and Y. J. Kim, “,e structural and functional organization of the yeast mediator complex,” Journal of Biological Chemistry, vol. 276, pp. 42003–42010, 2001. [14] F. J. Asturias, Y. W. Jiang, L. C. Myers, C. M. Gustafsson, and R. D. Kornberg, “Conserved structures of mediator and RNA polymerase II holoenzyme,” Science, vol. 283, no. 5404, pp. 985–987, 1999. International Journal of Journal of Peptides Nucleic Acids

The Scientifc International Journal of International Journal of World Journal Cell Biology Microbiology Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi http://www.hindawi.comwww.hindawi.com Volume 20182013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Anatomy Biochemistry Research International Research International

Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Submit your manuscripts at www.hindawi.com

Advances in Genetics Bioinformatics Research International Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Advances in International Journal of International Journal of Stem Cells BioMed Genomics Zoology International Research International Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 Virology www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Neuroscience Journal

Enzyme Journal of Journal of Research Parasitology Research Marine Biology Archaea Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018